SPONSORS:






User Tag List

Thanks Thanks:  0
Likes Likes:  0
Dislikes Dislikes:  0
Results 1 to 10 of 10
  1. #1
    Senior Member
    Join Date
    Oct 2006
    Location
    Vilnius, Lithuania
    Posts
    153
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Emulating active and inactive sessions separately

    Good day!

    Our company intends to do load testing of product (corporate tool) after migration to new technology (thin client on the Oracle Application Server Forms 10g). Performance of Oracle Application Server Forms 10g + client applet is area of interest.

    Objective is: for example it is 1000 concurrent users(are connected from start till end of the day); at one time 50 are active(doing things).

    Obvious approach is - we must emulate 1000 VU, and provide at any time at least 50 are active

    What our management intends to do:
    Make 50 active sessions(constantly doing things) + 950 inactive sessions(user is connected, maybe first form is loaded, makes no actions).

    What do you think about the last approach? Is it viable? What drawbacks has it?
    Roman

  2. #2
    Moderator JakeBrake's Avatar
    Join Date
    Dec 2000
    Location
    St. Louis - Year 2025
    Posts
    15,609
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    It sounds reasonable. Yet without knowing what the <font color="brown"> questions-to-be-answered-about-performance</font> are; one cannot say. On the inactive sessions, you must of course be aware and compensate for session timeouts. The best thing to do is understand the objectives of your management and how those objectives map to the <font color="brown"> questions-to-be-answered-about-performance</font>.

  3. #3
    Senior Member
    Join Date
    Oct 2006
    Location
    Vilnius, Lithuania
    Posts
    153
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    One of aspects of architecture of Oracle Applications Server Forms Services is: Runtime Area is created for each user.

    So variants are:

    Have 1000 runtime areas with small amount of actions in each

    or

    Have 50 permanently loaded runtime areas + 950 runtime areas with no actions

    What differences there can be?
    Roman

  4. #4
    Moderator JakeBrake's Avatar
    Join Date
    Dec 2000
    Location
    St. Louis - Year 2025
    Posts
    15,609
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    The differences and issues are far too voluminous to describe here. Again, the best thing to do is understand the objectives of your management and how those objectives map to the <font color="brown"> questions-to-be-answered-about-performance</font>.

    I suggest you browse some of the material at this link where you may find some relevant answers:
    http://www.sqaforums.com/showflat.ph...e=0#Post530183

  5. #5
    Senior Member
    Join Date
    Oct 2006
    Location
    Vilnius, Lithuania
    Posts
    153
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    trying to find information - there are plenty about web, but not so much about thin client with runtime areas...
    Roman

  6. #6
    Senior Member
    Join Date
    Oct 2006
    Location
    Vilnius, Lithuania
    Posts
    153
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    some additional information:

    Management's decision to emulate sessions separately was made without knowledge of system. And without proficience in performance testing. Pure theory


    Are there systems that behave differently when sessions are emulated separately and together? Maybe someone knows some other examples?
    Roman

  7. #7
    Moderator JakeBrake's Avatar
    Join Date
    Dec 2000
    Location
    St. Louis - Year 2025
    Posts
    15,609
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    [ QUOTE ]
    DeadGrunt: <font color="blue"> Management's decision to emulate sessions separately was made without knowledge of system.</font>

    [/ QUOTE ]
    That is fair in its own right. Hopefully other driving data included actual and or projected usage. I do not expect management to have a grasp on the specifics of modeling load. I expect that they are supportive and understand that performance testing is a project unto its own.
    .
    .
    [ QUOTE ]
    DeadGrunt: <font color="blue"> Management's decision to emulate sessions separately was made without knowledge of system. And without proficience in performance testing. Pure theory</font>

    [/ QUOTE ]Much is theory. Much is not.
    .
    .
    [ QUOTE ]
    DeadGrunt: <font color="blue"> Are there systems that behave differently when sessions are emulated separately and together? <font color="red">Yes. </font> Maybe someone knows some other examples? </font>

    [/ QUOTE ]Most every system will behave differently per those items you mentioned. The trick is to discover through specs, architectural and app design docs how a system is to behave. A simple example (per your concern) is a web/app/architecture that will expire your session after so much time. When a session is expired the system will behave differently in a number of ways. Among many other ways it will garbage collect your session out of the session bucket and perhaps perform better as a result. Again, this topic and all the various ways, means, methods, implementations, and technology behaviors are too vast to give you a comprehensive and robust answer.

  8. #8
    Senior Member
    Join Date
    Oct 2006
    Location
    Vilnius, Lithuania
    Posts
    153
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separately

    thanks for input, JakeBrake

    more precisely:
    Sessions in our app are not expiring at all, there is no session-timeout. Because it is a "main office app" and clients usually open sessions and work with it all day long
    Roman

  9. #9
    Senior Member
    Join Date
    Aug 2001
    Location
    Atlanta, GA
    Posts
    2,693
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separat

    I find that sometimes, if I'm not sure the way to most closely mimic or produce the load I'm looking for, I try it out both ways, on a smaller scale, to see what I might be missing.

    The first thing that comes to mind is whether there might be overhead involved in a user moving from an 'idle' to an 'active' state, and how much it will impact the system as a whole if there is a lot of that going on daily.

    To compare to a real life example. Perhaps at approximately 11:30EST, many of the EST users begin to become inactive, as they're going to lunch, but are still logged in. Perhaps when they all begin returning from 1-1:30 and activating their sessions again, there is a slowdown related to the overhead of those changes.

    I have no idea if that's the case in the scenario or DB you're using, as I've not worked with that platform - but it's certainly a question I'd want to prove out one way or the other before determining that I can accurately gauge system impacts or overall stability or performance by having 50 folks active with the same 950 folks inactive all day.

    If you feel management is operating solely on theory - take the time to run smaller-level tests to try and prove out your point. If there is nothing TO the point, then you've at least proven it to yourself as well and can focus on moving forward.
    Annemarie Martin
    Secretary
    Association for Software Testing

  10. #10
    Senior Member
    Join Date
    Oct 2006
    Location
    Vilnius, Lithuania
    Posts
    153
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Emulating active and inactive sessions separat

    thanks, QAGirl, for good ideas

    next morning we will have a meeting about performance testing strategy - and your suggestions helped me to understand processes, which can have impact on application

    about lunch:

    application clients are bank workers - most of them are "client service" - and they cannot go to lunch at one time. And now I think about back-office workers, who go to lunch at one fixed time - who can run some reports before going to lunch - and start working immediately when back to work after lunch...
    Roman

 

 

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Search Engine Optimisation provided by DragonByte SEO v2.0.36 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 12.50%
vBulletin Optimisation provided by vB Optimise v2.6.4 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging v3.2.8 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
vBNominate (Lite) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Feedback Buttons provided by Advanced Post Thanks / Like (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Username Changing provided by Username Change (Free) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
BetaSoft Inc.
Digital Point modules: Sphinx-based search
All times are GMT -8. The time now is 09:42 AM.

Copyright BetaSoft Inc.